Secondary Logo

Journal Logo

Original Investigation

Cognitive Testing and Exercise to Assess the Readiness to Return to Play After a Concussion

Sicard, Veronik1,2; Lortie, Jean-Christophe1; Moore, Robert Davis3; Ellemberg, Dave1,2

Author Information
Translational Journal of the ACSM: Summer 2020 - Volume 5 - Issue 11 - p 1-9
doi: 10.1249/TJX.0000000000000130



A concussion is a mild traumatic brain injury caused by biomechanical forces, including linear acceleration–deceleration and rotational forces acting on the brain (1). After sports-related concussion, athletes may experience several symptoms, including, but not limited to, headaches, confusion, dizziness, difficulty concentrating, and impairments in their oculomotor, vestibular, and cognitive functioning (1). These symptoms are believed to be clinical manifestations of complex cascades of neurometabolic, neurophysiological, and neurovascular changes (2).

In general, the clinical management of sports concussion rests on the recommendations of the consensus statement group and it is often paired with the assessment of cognitive functions after a period of cognitive and physical rest (1). The latest protocol, known as the Berlin return to play protocol (RTP), recommends slow increments in cognitive and physical activity at subsymptom threshold until the athlete has completed a full RTP. Specifically, once an athlete is asymptomatic at rest, a graduated return to sport is endorsed, starting with light exercise and progressing in a series of five steps to full-contact practice (1). If concussion-related symptoms arise during the progression, the athlete must return to a symptom-free status for at least 24 h and must go back to the previous step.

There is a critical point in the RTP protocol where an athlete progresses from noncontact training drills (step 4) to full-contact practice (step 5). The progression to step 5 places an athlete at a greater risk of incurring another concussion (2,3). Consequently, premature RTP is associated with a greater risk of subsequent injuries (including concussion), longer recovery time, and greater functional deficits. It also increases the risk of a more catastrophic condition known as second impact syndrome, where poor outcomes could lead to permanent disability or death (4). Thus, respecting the gradual RTP protocol reduces the risk of turning a transient brain injury into a chronic neurological condition.

Concussed athletes need medical clearance to progress to step 5. This important decision is complicated by the fact that the symptoms are nonspecific and subjective, that is athletes could voluntarily downplay or hide their symptoms to RTP faster or may not recognize that their symptoms result from a concussion (5). Given the low reliability of symptom reports, clinicians have come to rely more heavily on the cognitive clinical assessment to guide the safe RTP (1). The use of cognitive testing in conjunction with the appraisal of symptoms is more sensitive than relying on symptoms alone (6). Cognitive tests were believed to detect deficits or alterations after a concussion, which led to the development of several commercial computerized test batteries. Those batteries include tasks that are based on traditional neuropsychological measures. Some argue that computerized test batteries offer several advantages over traditional pencil-and-paper tests, including serial administration through alternative versions, improved response time measurement, ease of administration, and minimal time demand for the testing administrator (7). However, empirical evidence does not support the use of those computerized batteries for determining RTP. Existing batteries show low validity and reliability and do not seem sensitive to the consequences of concussion after the symptom resolution (7). This might have contributed to the belief that cognitive deficits resolve around the same time as symptoms do, that is, 14 to 21 d after concussion (8,9). Nevertheless, a growing body of literature suggests that long-term alterations may be observed using experimental cognitive paradigms several months to years after a concussion, and those alterations seem to be specific to the aspects of executive functions, such as cognitive flexibility, working memory, inhibition, and interference (for a meta-analysis, see Belanger et al. (10)). Thereby, using test batteries that do not include measures of executive functions may preclude the identification of concussed athletes who may not have fully recovered.

Another issue with the current RTP testing is the fact that the assessment is completed at rest. This is particularly important because participation in sports requires the ability to perform physically demanding exercises and simultaneously use perceptual information to make appropriate decisions, elaborate strategies, and inhibit surrounding distractors. In noninjured populations, an acute bout of moderate-to-vigorous aerobic exercise routinely induces cognitive enhancements, specifically in executive functions, as indicated by faster reaction time (RT) and increased response accuracy (ACC) (11–13).

A recent study found that the combination of cognitive testing and physical exercise might induce cognitive impairments in recently concussed athletes who were otherwise deemed to be ready to RTP (14). Specifically, McGrath et al. (14) conducted a retrospective review of medical records of 54 athletes who have suffered a concussion. Participants completed the ImPACT computerized test battery four times: (a) at preseason, (b) in the hours/days after concussion, (c) once asymptomatic and showing results similar to those of preseason, and (d) after an acute bout of aerobic exercise. Some athletes were assessed more than four times because they had to repeat the ImPACT battery several times before obtaining medical clearance to RTP (timepoint c). The athletes were separated into two groups, pass or fail, according to the results of the postexercise assessment (timepoint d) compared with their previous assessment (timepoint c). The authors found that one in four athletes (27.7%) who successfully completed step 4 of the RTP protocol and who had normal results on cognitive testing at rest (timepoint c) exhibited a reemergence of cognitive impairments after exercise. Although important, this study is not without methodological limitations. First, the exercise protocol was not standardized. That is, participants completed different exercise protocols (e.g., bike, treadmill, elliptical), at different intensities (60%–80% age-predicted maximal heart rate (APMHR)), and at different durations (range, 15–25 min) (14). Second, the sensitivity and reliability of the test battery used in the study (ImPACT) have often been criticized (15–17). Third, the study lacked a nonconcussed control group. Lastly, the order of testing (cognitive testing after rest vs cognitive testing after exercise) was not counterbalanced.

In general, the current procedure to determine readiness to RTP is based on self-reported symptoms and the assessment of cognitive functions with computerized test batteries. However, as previously explained, there are several limitations to this approach: 1) the test batteries show low-to-moderate validity and reliability, 2) the test batteries do not retain their sensitivity after symptom resolution, and most importantly, 3) the testing is completed at rest. Accordingly, the current study used a standardized exercise protocol to assess cognitive functioning before versus after physical exercise in asymptomatic university athletes who were cleared for full-contact practice (step 5 of the Berlin RTP protocol). Cognitive functioning was assessed by means of a switch task that was especially designed to measure executive functions. We expected that a subpopulation of athletes from the concussion group would show alterations with the combination of a standardized exercise protocol and a more sensitive cognitive task. In contrast, we expected an improvement in performance for controls, as an acute bout of moderate-to-vigorous exercise is known to induce cognitive enhancements in noninjured populations (11–13).



Eighty athletes from university rugby, soccer, football, hockey, volleyball, alpine skiing, and cheerleading teams took part in the study (see Table 1 for demographics). All participants were made aware of the purpose and procedures of this study, and they signed informed consent before participation. A structured interview was conducted at the beginning of the session with each participant to ensure that he/she met the inclusion criteria. Demographic and medical histories were obtained from each participant. Individuals with a history of neurological disease, psychiatric illness, non–sport-related head injury, learning disabilities, attention-deficit and hyperactivity disorder, and any other condition that may influence cognition were not invited to participate in the study. Furthermore, athletes who routinely used sedative medication, analgesic, or other central nervous system active medications, or who reported current or past use of illicit substances such as marijuana, stimulants, or opiates, or who reported drinking more than two glasses of wine, beer, or spirits per day were excluded from participation.

Participants Demographic Information.

Forty athletes who recently sustained a concussion were included in the concussion group. To control for variability in injury diagnosis and age at injury, only athletes incurring their concussion(s) during university sports were included in the concussion group. All concussions were identified by team medical staff and diagnosed within 24 h of injury by a physician from the university sports medicine clinic using the criteria established by the American Academy of Neurology (18) and the Consensus on Concussion in Sports (1). Athletes were referred to our research team by the medical staff once they successfully completed step 4 of the Zurich RTP protocol (at the time the study was conducted, the Berlin protocol was not yet published), but before returning to full-contact sports activities (step 5).

Forty teammates who never sustained a sport- or non–sport-related brain injury were included in the control group. To reduce the likelihood that an athlete with an undocumented concussion was placed in the control group, athletes in this group were also asked the following question: “Following a blow to the head, neck or body, have you ever experienced any concussion-like symptoms?” They were then presented with the list of clinical symptoms from the Sport Concussion Assessment Tool 3 (1). Anyone responding “yes” was excluded from participating in the study.


This study was completed in compliance with the Université de Montréal institutional research standards for human research. Individual testing took place in a quiet laboratory environment. After the semistructured interview, the completion of the Conners Adult ADHD Rating Scales and the symptom checklist from the Sport Concussion Assessment Tool 3, participants were randomly assigned to one of the eight subgroups. Each athlete had to complete the switch task twice during the same session: at rest and once after exercise. We counterbalanced the order of testing (rest vs postexercise condition) to minimize the effect of mental or physical fatigue stemming from cognitive and physical exercise. A 45-min resting period was completed before the cognitive testing in the rest condition. Furthermore, to reduce practice effects associated with repetitive testing, athletes completed alternate versions of the switch task. Half of each group completed version A first, whereas the other half completed version B first. At several times during the research protocol, athletes were asked if they were feeling any symptoms (i.e., after the cognitive testing, after the exercise protocol, at the end of their participation).

Exercise Protocol

The exercise condition consisted of 20 min of aerobic exercise on a cycle ergometer (Corival cpet, Lode, the Netherlands) at 80% of APMHR. This protocol was chosen as previous research indicates that it is associated with an improvement in cognitive function in healthy athletes, it is representative of the average intensity of physical exercise experienced by an athlete during a practice, and most importantly, it is the target heart rate (HR) corresponding to the one recommended by the Concussion in Sports Group at step 4 of the RTP protocol (1,19).

At the beginning of the exercise protocol, participants were asked to sit quietly for 5 min to record their resting HR using a Zephyr™ BioModule™ device (Medtronic, Annapolis, MD). Then, APMHR was calculated using Eq. (1) (20).


Participants began the protocol with a 2-min warm-up, pedaling at a frequency of 75–80 rpm with a load of 75 W. A modified Balke protocol was then used to reach 80% of the APMHR: resistance was gradually increased to reach the target HR within 5 min (21). Once 80% APMHR was reached, the 20-min exercise condition started. Whenever a participant’s HR dropped below 75% or exceeded 85%, resistance was adjusted to maintain HR on target. The exercise protocol ended with a 2-min cool-down.

Switch Task

The cognitive task was implemented in Psykinematix (version 1.5.1; KyberVision Japan LLC), and responses were recorded with a Cedrus serial-port response box (model RB-540; Cedrus, San Pedro, CA). For this task, stimuli were displayed on a black background in the center of the screen and consisted of circles or squares, either blue or green, with a white outline. Stimuli were 7 × 7 cm, with a visual angle of 4°. Stimuli were presented until the participant responded, with a maximum of 2000 ms allocated, and intertrial interval was 50 ms. In addition to the instructions on the monitor, the experimenter orally presented standardized instructions for each test block. Before each test block, the participants were reminded to respond as quickly and accurately as possible.

The switch task includes two conditions: homogeneous, which comprises two test blocks (color and shape), and heterogeneous. In the color test block, participants had to respond according to the color of the stimulus. Specifically, they were asked to press the left button if the stimulus presented was blue or the right button if the stimulus presented was green. After a 15-trial practice, participants completed 30 trials. In the shape test block, participants had to respond according to the shape of the stimulus. Specifically, they were asked to respond to the left if the stimulus was a circle or to the right if the stimulus was a square. Then, participants completed the heterogeneous condition during which they were asked to alternate their responses between the homogeneous rule sets (color vs shape) according to the outline (solid, dashed) of the stimuli. Thus, if the outline was solid, they were asked to respond according to the color rule set, and if the outline was dashed, they were asked to respond according to the shape rule set (see Fig. 1 for details). After a 30-trial practice, they completed three blocks of 60 trials. Across the three test blocks, participants completed a total of 90 switch trials and 90 nonswitch trials. Trials requiring a switch between stimulus-response rule sets were considered as switch trials, and trials that did not require a switch between stimulus-response rule sets were considered as nonswitch trials.

Figure 1
Figure 1:
Color–word switch task. The figure depicts the color–word switch task used in the current study. The call-out states the rule set that should be followed to elicit the appropriate response, which is written on the top right corner of each trial. For example, the first stimuli show a blue square with a dashed outline. Therefore, the participants must respond according to the shape, that is right button.

We developed a second version of the task (version B) because we wanted two alternative and equivalent versions for research testing. As we were concerned that the salience of the stimuli outline (solid line vs dashed line) might affect the RT, the difference between the two versions stands in the stimuli outline and the order of the two homogeneous conditions. In version B, participants start with the shape condition (stimuli have a solid outline) and then continue with the color condition (stimuli have a dashed outline).

Outcome Measures

Before calculating the average RT on the heterogeneous condition of the switch task (hetero RT), RT data were cleaned to remove inaccurate and outlying trials. That is, trials associated with an RT faster than 200 ms or slower than 2000 ms were excluded. Furthermore, only trials for which the responses were accurate were included in the averaged RT. Mean response ACC for the heterogeneous condition was computed (hetero ACC). Furthermore, we computed inverse efficiency scores (IES) because it has advantages over RT analyses when error rates are low (22). It was computed by dividing the mean RT by the mean ACC. Then, we computed the different switch costs (i.e., global RT, global ACC, global IES, local RT, local ACC, local IES, working memory RT, working memory ACC, working memory IES). Details on the computation of the switch costs are provided in Supplemental Digital Content 1,

Statistical Analysis

All statistical analyses were completed with SPSS 25.0 for Windows (IBM, Chicago, IL). Demographic information (age, height, weight, years of sports participation, years of education) and symptom information on the day of testing (number and intensity of symptoms) were analyzed by a series of independent-sample t-tests with the group as the between-subject factor. Moreover, exercise intensity and RPE during exercise were analyzed by a series of independent t-tests with the group as the between-subject factor. Given a sample size of 80 participants and a β of 0.02 (i.e., 80% statistical power), the current study had the requisite sensitivity to detect F-test interactions exceeding η2 ≥ 0.06 as computed using G*Power (Franz Faul, Universität Kiel, Germany). Output variables from the switch task (hetero ACC, hetero RT, hetero IES, and the different switch costs) were analyzed through a series of 2 × 2 × 2 × 2 (condition (rest, postexercise) × order of administration (rest–exercise, exercise–rest) × group (concussion, control) × version (A, B)) repeated-measures ANOVA. Pairwise comparisons using the Bonferroni correction were used to decompose significant interactions. Bivariate correlations were conducted between concussion information (i.e., number of prior concussions, symptoms duration, number and intensity of symptoms at time of injury, days since concussion) and results on the switch task. Correlations were adjusted for multiple comparisons by means of Bonferroni corrections.

Based on their results on the switch task during both conditions (rest, postexercise), athletes from the concussion group were categorized into the pass or fail group for each outcome measure. Specifically, they were placed in the fail group if their score was 2 SD lower than the control group’s average score. This criterion was chosen since 95.45% of the values lie within 2 SD of the mean; therefore, a score that falls outside of the two-sigma effect (95%) may be considered as an outlier. Importantly, no outlier was found in the control group, allowing us to use this method with confidence. χ2 analyses were conducted to test for equality of proportions of failure between conditions for each variable.


Data are presented as mean ± SD.

Demographic Information and Concussion History

Demographic information can be found in Table 1. No group difference was observed for age, height, weight, years of sports participation, and years of education (t(78) ≤ 1.78, P ≥ 0.08). The number of symptoms experienced on the day of testing did not differ between groups (t(78) = 0.99, P = 0.33), nor did the intensity of symptoms (t(78) = 1.34, P = 0.18). None of the athletes reported a change in symptoms after cognitive testing or the exercise protocol. Within the concussion group, the average number of prior concussions was 1.8 ± 1.1 (range, 0–4). Athletes participated in this study on an average of 11.2 ± 7.5 d after concussion (range, 6–32 d). The average number of symptoms reported at the time of injury was 11.2 ± 6.9 (range, 2–29), and the average intensity of those symptoms was 31.1 ± 27.23 (range, 3–107). No athlete suffered from a loss of consciousness and posttraumatic amnesia. Symptoms lasted an average of 6.82 ± 6.78 d (range, 1–26 d).


Details concerning the parameters of the physical exercise condition are presented in Table 2. No difference was observed between the concussion and control groups for resting HR (t(78) = 0.29, P = 0.77). Furthermore, no group difference was observed in power (in watts), RPE, average HR, and %APMHR for the exercise condition (t(78) ≤ 1.44, P ≥ 0.16).

Exercise Parameters.

Group Performance on the Switch Task

Results are presented in Fig. 2. None of the interactions were significant (F(1,76) ≤ 4.19, P ≥ 0.05, η2 ≤ 0.05, φ ≤ 0.52). Main effects of group were found for hetero ACC, hetero IES, global ACC cost, global IES cost, working memory ACC cost, and working memory IES cost (F(1,76) ≥ 4.73, P ≤ 0.03, η2 ≥ 0.06, φ ≥ 0.50). Specifically, athletes in the concussion group exhibited a lower hetero ACC (86.2% ± 14.8%) and a higher IES (11.8 ms/% ± 3.3ms/%) relative to those in the control group (93.0% ms/% ± 3.9%; 9.5 ± 1.63ms/%; t(78) ≥ 2.18, P ≤ 0.03), irrespective of condition (rest vs postexercise). Furthermore, the concussion group exhibited a higher global ACC cost (11.2% ± 14.4%) and a higher global IES cost (6.6 ms/% ± 3.2ms/%) relative to the control group (4.1% ± 3.6%; 5.4 ms/% ± 1.6ms/%; t(78) ≥ 2.24, P ≤ 0.03). Finally, athletes in the concussion group exhibited a higher working memory ACC cost (9.6% ± 14.7%) and a higher working memory IES cost (5.9% ± 3.2%) relative to controls (2.8% ± 3.7%; 4.64 ± 1.3; t(78) ≥ 2.27, P ≤ 0.03). No main effect of condition, order of administration, or version of the task was observed for any variables (F(1,76) ≤ 3.28, P ≥ 0.07, η2 ≤ 0.04, φ ≤ 0.43). A detailed presentation of results is presented in Supplemental Digital Content 2,

Figure 2
Figure 2:
Performance on the switch task. The figure depicts significant group performance on the switch task on both conditions (rest, postexercise) for hetero ACC (A), hetero IES (B), global ACC cost (C), global IES cost (D), working memory ACC cost (E), and working memory IES cost (F). *Difference at P ≤ 0.05. **Difference at P ≤ 0.01. ***Difference at P ≤ 0.001.


Bivariate correlations were conducted between significant variables (i.e., hetero ACC, hetero IES, global ACC cost, global IES cost, working memory ACC cost, and working memory IES cost) and concussion characteristics. A significant relationship was observed between the intensity of symptoms at time of injury and working memory ACC cost (r = 0.33, P = 0.04); however, it does not remain significant after statistical correction for multiple comparisons. No other significant relationship was observed (r ≤ 0.32, P ≥ 0.05).

Individual Scores at Rest and Postexercise

χ2 Analyses were conducted to determine whether the group differences could be attributed to a subgroup of concussed athletes. Irrespective of condition (rest or postexercise), up to 30.0% of athletes (12/40) from the concussion group were 2 SD below the control group’s average score. Specifically, of the total number of athletes with a concussion, based on their ACC, 2.5% (1/40) failed only at rest, 10% (4/40) failed only at postexercise, and 17.5% (17/40) failed on both conditions. Distribution of data is presented graphically in Fig. 3. χ2 Analyses indicated no significant difference in proportions of failure between the rest and postexercise conditions for the concussion group (χ2(1,40) ≤ 2.18, P ≥ 0.14). Complete results are presented in Table 3. A detailed presentation of results is presented in Supplemental Digital Content 2,

Figure 3
Figure 3:
Distribution of the performance on the switch task. The figure depicts the distribution of the performance on the switch task of the concussion group for hetero ACC at rest (A), hetero ACC at postexercise (B), hetero IES rest (C), hetero IES postexercise (D), global ACC cost at rest (E), global ACC cost at postexercise (F), global IES cost at rest (G), global IES cost at postexercise (H), working memory ACC cost at rest (I), working memory ACC cost at postexercise (J), working memory IES rest (K), and working memory IES postexercise (L). The dashed line represents the control group’s average score. The solid line represented the 2 SD cutoff used to determine the fail or pass status. For ACC, the scores below the solid line were considered as failure, whereas for IES, those above the solid line were considered as failure.
Description of Total Participants, Passes, Failures, and χ2 Analysis.


This study aimed to examine preexercise and postexercise cognitive performance among student-athletes who were asymptomatic and cleared to RTP after a sport-related concussion. Two main results emerge. First, despite reporting being asymptomatic, athletes in the concussion group exhibited a lower cognitive performance relative to the control group. Second, irrespective of condition (rest or postexercise), up to 30% of recently concussed athletes who successfully completed step 4 of Zurich RTP protocol exhibited cognitive impairments relative to their teammate controls.

Despite reporting being asymptomatic, athletes in the concussion group exhibited a lower cognitive performance, as evidenced by low ACC, higher inverse efficiency, and higher global and working memory costs on the switch task relative to the control group. These results are in line with a growing body of literature suggesting significant group differences that persist beyond clinical recovery (23–30), where clinical recovery is defined functionally as a return to normal activities, including school and sports, after injury (31). In the current study, a subgroup of athletes who were cleared to RTP after their concussion exhibited diminished performance in multiple aspects of executive functions relative to teammate controls, suggesting that although clinical recovery was achieved, neurophysiological recovery was not completed.

Most importantly, up to 30% of recently concussed athletes who successfully completed step 4 of Zurich RTP protocol exhibited cognitive impairments relative to their teammate controls, irrespective of condition (rest, postexercise). These results are similar to previously published data (14), but expand on prior findings by revealing more multifaceted deficits (in ACC and RT) across multiple cognitive domains (working memory, cognitive control, and cognitive strategy). By contrast to other cognitive tasks routinely used for tracking recovery after a concussion, the switch task requires active, top-down regulation to facilitate successful performance. Hence, it may be more sensitive to the cognitive dysfunction after a concussion as indicated by the larger differences in performance between the concussion group and the control group in the current study. Although no main effect of exercise was observed, our data support the use of postexercise cognitive testing to assess the readiness of an athlete to RTP, because 10% of athletes only showed deficits after exercise.

Clinical Implications

Determining when an athlete is ready to RTP after a concussion is a critical decision for sports medicine professionals. Although medical personnel commonly monitor athletes for the reemergence of symptoms after exercise, they rarely carry out postexercise cognitive testing. Relying on clinical symptoms to guide decision-making is not optimal as many athletes may withhold their symptoms for several reasons including, but not limited to, the desire to stay in the game, the fear of letting down their teammates, or the inability to recognize their symptoms (5). Furthermore, many studies indicate that athletes can still exhibit cognitive dysfunctions even when deemed to be “asymptomatic.”

Similar to prior studies, we found cognitive deficits continuing after the disappearance of symptoms (14). By going beyond basic cognitive tasks provided by prepackaged batteries, we revealed a multifaceted pattern of cognitive dysfunction, which included aspects of working memory, attention, and cognitive control. Therefore, the current findings further reinforce the need for more sophisticated tests of complex cognitive processes to reveal deficits that persist after a concussion. In addition, our results suggest that physical exercise improves the identification of athletes with residual cognitive deficits when used in conjunction with cognitive testing. Our results are of important clinical relevance as they reveal that 10% of athletes who passed cognitive testing at rest failed after exercise. Thus, some asymptomatic athletes who are deemed ready to RTP will experience a reemergence of cognitive dysfunction after moderate-to-vigorous exercise. This is of critical importance as it may place them at risk of subsequent injury and more serious cognitive impairment.

Our findings also highlight the importance of considering interindividual differences in recovery trajectories after concussion. Indeed, although many recently concussed athletes show normal cognitive functioning at rest and postexercise, 30% of them performed exhibited cognitive deficits. Furthermore, as indicated previously, 10% of those athletes only exhibited those deficits after exercise. Thus, critical information on the cognitive recovery of individual patients may be lost by averaging group results. Clinicians and researchers should be cognizant of this interindividual variability when interpreting research findings as well as implementing preexercise and postexercise cognitive assessments to their clinical practices.


The results should be interpreted considering some methodological limits. First, the current sample was limited to university student-athletes, and therefore, the findings are not generalizable to other age groups. Nevertheless, this allowed us to strictly control the diagnostic criteria, adherence to the Zurich RTP, and consistency of testing. Moreover, the cross-sectional nature of this study can also be viewed as a limit, because it does not preclude the possibility that some preexisting differences contributed to the current observations. To minimize this possibility, athletes from concussion and control groups were carefully matched for age, level of education, body mass index, sport, and position, and no participant had a history of neurodevelopmental or neurological disorders. Second, identifying and specifying the length of time each athlete took during and in between the steps of the RTP protocol would have been pertinent information. It is possible that a longer interval between some steps may be predictive of slower recovery. Future studies should record and include this information, as this will provide clinicians with guidelines on the typical length of time to expect for each step. Third, the fact that athletes completed the cognitive test twice during the same session can be viewed as a limit due to the possible lingering effect of exercise on cognition and practice effects associated with repeated testing on such a short interval. However, this possibility is minimized given that the conditions were counterbalanced. Fourth, the current study only examined the effect of an acute bout of aerobic exercise on cognition, posing a threat to the ecological validity of the current research protocol, as athletes likely perform a combination of aerobic, anaerobic, and dynamic exercises when they are practicing their sports. Finally, only executive functions were assessed in the current study, limiting the generalizability of the results to other facets of brain functioning and outcomes of concussion. A comprehensive assessment would be necessary to best identify athletes who have not recovered from their concussion (e.g., cerebral blood flow, neuroelectric function, motor, vestibular, and ocular-motor function).


The current study suggests that more than one out of four athletes who successfully completed the RTP protocol still exhibited diminished cognitive functioning compared with controls. Furthermore, 10% of athletes normally performed at rest, yet exhibited a reemergence of cognitive dysfunction after exercise. Thus, the use of a sensitive cognitive task combined with physical exercise should be used to assess readiness to RTP.

We would like to thank Alexe Simard, Alejandra Martinez, Annik Charest-Rettig, Ariane Cormier, Catherine Guimond, Florence Boutet, Najd Ajji, and Olivier Cyr for their help in data collection and data entry.

The authors have no conflict of interest to declare. They have no professional relationships with companies or manufacturers who would benefit from the results of the present study. This research was supported by a grant from the Canadian Institutes of Health Research (CIHR MOP-102608) to D. E. The results of the present study do not constitute an endorsement by the American College of Sports Medicine.


1. McCrory P, Meeuwisse W, Dvorak J, et al. Consensus statement on concussion in sport-the 5th international conference on concussion in sport held in Berlin, October 2016. Br J Sports Med. 2017;51:838–47.
2. Giza CC, Hovda DA. The new neurometabolic cascade of concussion. Neurosurgery. 2014;75(Suppl 4):S24–33.
3. Laurer HL, Bareyre FM, Lee VM, et al. Mild head injury increasing the brain's vulnerability to a second concussive impact. J Neurosurg. 2001;95(5):859–70.
4. McCrory P, Davis G, Makdissi M. Second impact syndrome or cerebral swelling after sporting head injury. Curr Sports Med Rep. 2012;11(1):21–3.
5. Delaney JS, Caron JG, Correa JA, Bloom GA. Why professional football players chose not to reveal their concussion symptoms during a practice or game. Clin J Sport Med. 2018;28(1):1–12.
6. Van Kampen DA, Lovell MR, Pardini JE, Collins MW, Fu FH. The “value added” of neurocognitive testing after sports-related concussion. Am J Sports Med. 2006;34(10):1630–5.
7. Mayers LB, Redick TS. Clinical utility of ImPACT assessment for postconcussion return-to-play counseling: psychometric issues. J Clin Exp Neuropsychol. 2012;34(3):235–42.
8. Bleiberg J, Cernich AN, Cameron K, et al. Duration of cognitive impairment after sports concussion. Neurosurgery. 2004;54(5):1073–8; discussion 1078-1080.
9. Iverson GL, Brooks BL, Collins MW, Lovell MR. Tracking neuropsychological recovery following concussion in sport. Brain Inj. 2006;20(3):245–52.
10. Belanger HG, Spiegel E, Vanderploeg RD. Neuropsychological performance following a history of multiple self-reported concussions: a meta-analysis. J Int Neuropsychol Soc. 2010;16(2):262–7.
11. Chang YK, Labban JD, Gapin JI, Etnier JL. The effects of acute exercise on cognitive performance: a meta-analysis. Brain Res. 2012;1453:87–101.
12. Lambourne K, Audiffren M, Tomporowski PD. Effects of acute exercise on sensory and executive processing tasks. Med Sci Sports Exerc. 2010;42(7):1396–402.
13. Smith PJ, Blumenthal JA, Hoffman BM, et al. Aerobic exercise and neurocognitive performance: a meta-analytic review of randomized controlled trials. Psychosom Med. 2010;72(3):239–52.
14. McGrath N, Dinn WM, Collins MW, Lovell MR, Elbin RJ, Kontos AP. Post-exertion neurocognitive test failure among student-athletes following concussion. Brain Inj. 2013;27(1):103–13.
15. Alsalaheen B, Stockdale K, Pechumer D, Broglio SP. Validity of the Immediate Post Concussion Assessment and Cognitive Testing (ImPACT). Sports Med. 2016;46(10):1487–501.
16. Resch JE, McCrea MA, Cullum CM. Computerized neurocognitive testing in the management of sport-related concussion: an update. Neuropsychol Rev. 2013;23(4):335–49.
17. Broglio SP, Ferrara MS, Macciocchi SN, Baumgartner TA, Elliott R. Test–retest reliability of computerized concussion assessment programs. J Athl Train. 2007;42(4):509–14.
18. American Academy of Neurology. Practice parameter: the management of concussion in sports (summary statement). Report of the Quality Standards Subcommittee. Neurology. 1997;48(3):581–5.
19. Moore RD, Romine MW, O'Connor PJ, Tomporowski PD. The influence of exercise-induced fatigue on cognitive function. J Sports Sci. 2012;30(9):841–50.
20. Gellish RL, Goslin BR, Olson RE, McDonald A, Russi GD, Moudgil VK. Longitudinal modeling of the relationship between age and maximal heart rate. Med Sci Sports Exerc. 2007;39(5):822–9.
21. Hanson P. In: Strauss R, editor. Clinical Exercise Training. Philadelphia (PA): W.B. Saunders Company; 1984.
22. Hughes MM, Linck JA, Bowles AR, Koeth JT, Bunting MF. Alternatives to switch-cost scoring in the task-switching paradigm: their reliability and increased validity. Behav Res Methods. 2014;46(3):702–21.
23. Sicard V, Moore RD, Ellemberg D. Sensitivity of the cogstate test battery for detecting prolonged cognitive alterations stemming from sport-related concussions. Clin J Sport Med. 2019;29:62–8.
24. Fazio VC, Lovell MR, Pardini JE, Collins MW. The relation between post concussion symptoms and neurocognitive performance in concussed athletes. NeuroRehabilitation. 2007;22(3):207–16.
25. Moore RD, Pindus DM, Drolette ES, Scudder MR, Raine LB, Hillman CH. The persistent influence of pediatric concussion on attention and cognitive control during flanker performance. Biol Psychol. 2015;109:93–102.
26. Ellemberg D, Leclerc S, Couture S, Daigle C. Prolonged neuropsychological impairments following a first concussion in female university soccer athletes. Clin J Sport Med. 2007;17(5):369–74.
27. Broglio SP, Pontifex MB, O'Connor P, Hillman CH. The persistent effects of concussion on neuroelectric indices of attention. J Neurotrauma. 2009;26(9):1463–70.
28. Howell D, Osternig L, Van Donkelaar P, Mayr U, Chou LS. Effects of concussion on attention and executive function in adolescents. Med Sci Sports Exerc. 2013;45(6):1030–7.
29. Mayr U, LaRoux C, Rolheiser T, Osternig L, Chou LS, van Donkelaar P. Executive dysfunction assessed with a task-switching task following concussion. PLoS One. 2014;9(3):e91379.
30. Moore RD, Hillman CH, Broglio SP. The persistent influence of concussive injuries on cognitive control and neuroelectric function. J Athl Train. 2014;49(1):24–35.
31. Iverson GL, Gardner AJ, Terry DP, et al. Predictors of clinical recovery from concussion: a systematic review. Br J Sports Med. 2017;51(12):941–8.

Supplemental Digital Content

Copyright © 2020 by the American College of Sports Medicine